MIXING INCOME GROUPS IN PUBLIC HOUSING
by
Brent W. Ambrose
and
William Grigsby
Working Paper #354
2
FINAL [3050 wds]
MIXING INCOME GROUPS IN PUBLIC HOUSING
Brent W. Ambrose, William Grigsby
Federal low-income housing policy has shifted markedly over the last decade. The 1986 Low
Income Housing Tax Credit program, which made households with incomes of up to 60 percent
of area median income eligible for rental assistance, has expanded assistance to include the
working poor. By 1993, growing frustration over public housing conditions fueled the creation
of another new federal housing assistance program--HOPE VI--intended, according to the
Department of Housing and Urban Development, "to end the physical, social and economic
isolation of obsolete and severely distressed public housing by recreating and supporting
sustainable communities and lifting residents from dependence and persistent poverty,"
Hope VI requires participating public housing authorities (PHAs) to form public/private
partnerships in order to leverage funding from other programs. The linkage between leveraged
finance and mixed-income development is a key element of HOPE VI. Leveraged finance
combines public inputs with private development professionals, while shifting responsibility for
assisted housing from the federal to the local level.
2
3
Perhaps the most significant impact of the public housing leveraged finance program is
improved asset management. The goal is to de-institutionalize PHA properties by including
higher income residents and commercial space, hence improving product quality without
increasing subsidies. Over the past five years HOPE VI has received national attention. Since its
$300 million initial authorization, the program has expanded to a FY 1999 appropriation of
$550 million. HOPE VI grants are awarded to local PHAs through a competitive process. A
minimum of 80 percent of the funds received must be allocated to capital costs, including
reconstruction, site improvement, and construction of replacement units. A maximum of 20
percent of the grant can be used for community and support service programs such as job
training, literacy programs, daycare, and youth activities. In addition, the city is obliged to match
15 percent of the support-services allocation with non-federal contributions. Local PHAs have
wide latitude in the use of HOPE VI funds. For example, the Cleveland, Ohio housing authority
is using HOPE VI funds to create a “social service mall” in a mid-rise office building;
Milwaukee is creating “micro neighborhoods” by demolishing deteriorated housing and
constructing connecting streets.
The Atlanta Housing Authority has taken the lead in combining its HOPE VI funds with
private financing to leverage its development activities. Several aspects of the Atlanta experience
provide useful insights into how mixed-income projects can affect the isolation of low-income
groups in a large metropolitan area.
3
4
THE ATLANTA EXPERIENCE
As the eighth largest public housing authority in the United States, Atlanta's Housing Authority
(AHA) is responsible for roughly 13 percent of the city’s housing stock, comprising more than
14,000 public housing units and more than 7,000 privately-owned rental units subsidized
through the Section 8 housing allowance program. The average income of residents in AHA
public housing units is $6,500. One-third of the residents are on welfare, and one-third are older
than 65.
Capitalizing on the publicity and development activity associated with the 1996
Olympics, the AHA created the $300 million Olympic Legacy Program. The goals of the
Program are to: assist public housing residents in achieving self-sufficiency through education,
job training and employment; leverage federal resources in order to attract private capital and
other state and local funds; assist the AHA in generating an unrestricted income stream from
development activities; and, use AHA programs to revitalize neighborhoods. AHA's aim is to
demolish approximately one-third of its public housing units, replacing them with mixed-income
developments and Section 8 housing. Two of its public housing projects--Techwood/Clark
Howell and East Lake Meadows--have already been partially reconstructed and rented to new
families.
4
5
Techwood Homes, a development of 1,081 units in downtown Atlanta, was America's
first public housing project. By the mid-1980s, poor design, inadequate resident screening, lack
of maintenance, and management neglect resulted in the deterioration of the housing project into
a crime infested urban ghetto. Using HOPE VI funds, AHA is replacing the original superblock
design with Centennial Place, a 900-unit mixed-income project. The housing units are grouped
in small clusters, and the original street grid has been rebuilt using the original street names. The
clear delineation of public and private spaces, parking behind housing units, and traditional
architectural styles, are designed to create an environment that is similar to high quality
residential projects.
The developers of Centennial Place originally proposed an equal mix of market-rate
units, Low Income Housing Tax Credit (LIHTC) units (aimed at the working poor), and public
housing units. After negotiations between AHA and the resident association, 40 percent of the
units are market-rate, 40 percent are public housing, and 20 percent are LIHTC. The public
housing units are not pre-designated, and all units are similar in design.
This mix resulted from the fear of many original residents that that they would become
homeless if new social programs moved them off welfare and into income groups not served by
public housing. Introducing LIHTC units into the mix assured residents that even if their incomes
rose above public housing limits, they could remain in Centennial Place.
Rental rates and income limits for the subsidized units at Centennial Place follow HUD
and AHA guidelines. HUD requires that the maximum income level for persons living in public
5
6
housing units is 80 percent of area median income (AMI). The agreement between AHA and
the private developer specifies that half of the public housing families at Centennial Place have
incomes of less than 30 percent of AMI, and half may have income greater than 30 percent of
AMI. With the median resident income set at this level, it is calculated that rental income will
cover the cost of operation, and no operating subsidies will be required.
Rents for market-rate and tax-credit units are set independently of the income levels of
public housing eligible families. The agreement further specifies that while the private developer
financially benefits from the tax-credit and market-rate units, he may not benefit from the public
housing units. The income and rental requirements are further complicated using LIHTC to
finance 20 percent of the units. LIHTC guidelines specify that rents for LIHTC units cannot
exceed 60 percent of AMI. However, LIHTC rules do not require accepting residents with
incomes below 30 percent of AMI. Consequently, the average LIHTC resident will probably
have a higher income than the average public housing resident.
As of March 1998, 358 units in Phase I and II were leased (Table I). Surprisingly, the
market-rate and LIHTC units have leased faster than the public housing units. The current ethnic
mix is 70 percent African-American, 20 percent Caucasian, and 10 percent Hispanic and
Asian.
6
7
Table I. Centennial Place, Phase I
Monthly
Market Rent
1 BR apartment
2 BR apartment
3 BR apartment
2 BR townhouse
3 BR townhouse
4 BR townhouse
Total
$679
$779
$879
$899-$999
$1259-$1359
$1499
Market
Rate
31
6
24
6
5
2
74
No. of units
LIHTC
Public
Housing
24
11
4
23
0
0
4
17
4
18
0
2
36
71
Several neighborhood facilities have also been constructed, including: a new elementary
school; a community center; corporate suites; a shopping center with a 40,000 square-foot
supermarket; a Holiday Inn Express (on an adjacent site); a new YMCA; a police miniprecinct; and a branch bank. The presence of these facilities, combined with the quality of the
housing, has enabled the developers to attract market-rate residents.
East Lake Meadows, located in the southeast section of Atlanta, represents a truly
unique case of public housing revitalization. An affluent community in the 1920s, the East Lake
neighborhood deteriorated in the 1960s. In 1968, the once elite East Lake Golf Club was sold
as the original members moved to the suburbs. In 1993 the Cousins Foundation bought the East
Lake Golf Club with the purpose of revitalizing the adjacent public housing complex while
paying tribute to the legendary Bobby Jones, who learned to play golf on the course. Club
members have committed $8 million to revitalization efforts, and income from corporate
7
8
memberships is being used to fund the East Lake Community Foundation (ELCF), which serves
as the non-profit developer of the new mixed-income community. Working with AHA, ELCF is
demolishing 650 public housing units and replacing them with a 550-unit complex of
townhouses, garden style apartments, and duplexes. The new units are being placed around a
new public golf course financed, developed and operated by ELCF.
Although no HOPE VI funds are used in this development, AHA has used Hope VI as
its model. Like Centennial Place, the design for East Lake Meadows is the result of negotiations
between AHA, the ELCF, and residents. The project is evenly divided between market-rate
and public-housing units (Table II).
Table II. East Lake Meadows -- Phase I
2 BR apartment
3 BR apartment
2 BR townhouse
3 BR townhouse
4 BR townhouse
Monthly
Market
Rent
$ 830
$ 925
$ 840
$ 975
$1,050
Number of Units
Market
Public Housing
Rate
20
20
15
15
25
26
25
26
5
5
90
92
Both Centennial Place and East Lake Meadows rely on a public/private financing
partnership. Centennial Place follows the leveraged finance concept of the HOPE VI, with the
private developer entering into a 55-year ground lease with AHA. HOPE VI funds are
8
9
provided in the form of a long-term loan that is repaid to AHA from rental income. AHA also
receives 25 percent of the development fee, plus one-third of the annual net cash flow over the
55-year period. Finally, 20 percent of Centennial Place is financed through the sale of low
income housing tax credits.
East Lake Meadows has a private developer, with the AHA being granted permission
by HUD to contribute federal “Development Funds” rather than HOPE VI grant funds. ELHC
uses the income generated from the market-rate units to fund Foundation activities at the site.
As at Centennial Place, the developer holds a 55-year ground lease and all funds contributed by
non-AHA sources are used to fund the non-public housing units and project amenities, including
the public golf course and a community center.
REASONS FOR SUCCESS
To date, Centennial Place has succeeded in attracting different income groups. Phases I and II
are fully leased, and strong leasing continues. East Lake Meadows, although not as far along in
the development process, has also successfully rented both market-rate and public housing
units. Significantly, Centennial Place and East Lake Meadows have attracted market-rate
residents without offering rent concessions. In fact, rents for Centennial Place are actually higher
than rents in similar Atlanta rental properties, while rents at East Lake Meadows are slightly
lower (Table III).
9
10
Table III: Rent and Expense Comparisons of Hope VI and Class A Apartments in Atlanta
Centennial Place
East Lake
Average Class A
Average
Rent/SF
$0.89
$0.69
$0.76
Annual Operating
Expenses/Unit
$3,600
$3,500
$3,700
Construction
Costs/Unit
$73-78,000
$70,000
not available
The rental success of these two projects--and of other Hope VI projects around the
country--has surprised many analysts who felt that it would be difficult to find market-rate
renters willing to live in developments with a high proportion of low-income tenants. There are
three keys to the initial success of these projects. First, both are located in attractive, highly
accessible neighborhoods. Second, both are well designed and visually appealing projects that
do not look like “public housing projects” but rather “Class A” rental housing, and offer the
same amenities as market projects (front gates, club houses, security systems, etc.). Finally, they
are well managed, particularly with respect to resident screening, insuring that only applicants
with middle-class behavior patterns are accepted.
BARRIERS TO EXPANSION
10
11
Given the initial success of the Atlanta projects, the question arises whether mixed-income
housing developments should be a major part of a city’s subsidized housing stock. The AHA
plans to convert 4,800 public housing units, or one-third of its inventory, into similar
developments. To achieve this goal four major barriers must be overcome. First, a sufficient
number of sites must be found that can attract market-rate residents. Assuming the 50/50 split
of market-rate residents to public housing residents at East Lake Meadows is feasible for a
successful mixed-income community, then AHA’s goal of converting over 4,800 public housing
units will require 4,800 market rate units (including the units currently built at Centennial Place
and East Lake Meadows). If the size of the average development is 500 units, then the AHA
goal implies that 19 conversion sites must be found. Because AHA is already in the process of
converting 7 projects (in addition to Centennial Place and East Lake Village) into mixed income
developments, it would need to find at least 10 other project sites for conversion in order to
meet its goal. Furthermore, it is unclear whether there is sufficient demand for an additional
4,800 units of market-rate housing.
The nine currently completed and planned mixed-income projects are in pleasant
neighborhoods, with natural buffers separating them from the negative effects of nearby poverty
areas. They also have easy access to local employment, cultural and entertainment. Finding
suitable sites may be difficult. Attractive privately owned sites will be prohibitively expensive,
while most public-housing sites are in unattractive locations.
11
12
A second barrier is finding management teams for these projects. Management
personnel at Centennial Place and East Lake Meadows are highly motivated, in part because of
the pioneering nature of their efforts. Cloning such persons for a large number of subsequent
projects is likely to be difficult. However, the integration with private sector housing does
provide a deeper talent pool than does public housing.
The third barrier concerns operating costs. The rental income from subsidized residents
must be sufficient to keep the per-family operating public subsidies minimal. The agreement at
Centennial Place targets an average income of the public housing residents at 30 percent of
AMI (or $16,410 in Atlanta), with 50 percent above $16,410 and 50 percent below $16,410.
This effectively ensures that the rental income from the public housing units covers their
operating costs. The conversion of 4,800 public housing units to mixed-income developments
means that 2,400 units (equivalent to 16% of the current public housing stock) will have to be
successfully rented to families with incomes above $16,410. However, the current average
income for APA residents is a mere $6,500, with fully 37 percent of the residents having
incomes below $5,000 and 95 percent having incomes below $15,000. Given this resident
profile, the goal of renting 16 percent of the public housing stock to families above 30 percent of
AMI may be unrealistic.
In order to minimize project subsidies, AHA has an incentive to rent as many units as
possible to residents with incomes well above those of current AHA clientele. Similar incentives
also exist with respect to all HOPE VI projects, as the cost of new construction or substantial
12
13
rehabilitation dictates similar rent levels. HUD regulations provide a way for AHA to respond to
this incentive to be selective while still abiding by the first-come/first-serve rule that requires
AHA to accept qualified applicants in the order in which they apply, without regard to their
income. Although the regulations prevent AHA from discriminating against a prospective
resident on the basis of income, they do allow local housing authorities to give preference to
eligible applicants who have a working member in the household. On average, such households
have higher incomes than the low-income applicant pool as a whole.
The final barrier is the large capital cost of new or substantially rehabilitated housing
relative to the incomes of the subsidized occupants. Construction costs at Centennial Place and
East Lake Meadows averaged more than $70,000 per unit. Development fees and construction
financing probably added at least another $10,000 per unit (land was "free"). Amortizing this
$80,000 capital investment at 7 percent over 30 years yields an annual cost of about $6,400, a
full 40 percent of the average targeted public-housing occupant income of $16,410. This cost is
borne entirely by the taxpayer, since it is not offset by rental income, all of which is used to
cover operating costs. Such a sizeable subsidy might not be politically acceptable on a large
scale. And even if AHA achieves its ambitious income-mixing objective, most of its tenants-and most low-income Atlantans--will remain in low-income neighborhoods.
Do HOPE VI projects reduce the residential isolation of the poor? Centennial Place
obviously represents a greater range of incomes than the Techwood project it replaced. But it is
unclear whether the 1,081 former Techwood residents, only 64 of whom returned to Centennial
13
14
Place, are less isolated than before being displaced, and whether the 540 new subsidized
households, who on average have higher incomes than those who were displaced, are less
isolated than they were prior to moving to Centennial Place. Equally important, we do not
know whether the arrival of the former Techwood/Centennial residents in their new
neighborhoods and the departure of the new Centennial residents from their old neighborhoods
increased or decreased income mixing in those areas. In the absence of this information, we
cannot conclude that income mixing at the project level produced a net increase in income
mixing in the Atlanta housing market overall.
Nevertheless, assuming that the initial success of projects like Centennial Place and East
Lake Meadows is not temporary, then HOPE VI has performed a truly remarkable service in
dispelling the entrenched belief that it is not possible to design financially successful projects in
which there is income-mixing a the block level. If the families in these projects have similar
values even though their incomes are quite different, then income integration seems possible, at
least in the rental sector.
[Author bio: Brent W. Ambrose is Associate Professor of Finance and Director of the
Center for Real Estate Research at the University of Kentucky Gatton School of
Business Administration.]
[Author’s bio: William Grigsby is Professor Emeritus of City and Regional Planning at
the University of Pennsylvania.]
14
15
[Endnote: Kyla Weissman, Melissa Lober, and Alexandra Deas assisted in this research.
Financial support for this research project provided by the J. Ronald Terwilliger Fund for
Housing Research at the Samuel Zell and Robert Lurie Real Estate Center at Wharton.]
15